优享资讯 | 我开源了一个通用的数据翻译框架easy-trans

掘金后端 ( ) • 2024-04-25 17:16

一、背景

为啥会有这个项目呢？我在研发项目中，经常有需要进行数据翻译的地方，比如需要把几张表的关联查询结果返回给前端，现在用的都是自动化的orm框架，大概逻辑如下：

// 查询a表数据
List<A> alist = aDao.findAll(query);
// 根据a表数据，获取b表id
List<Long> bids = alist.stream().map(a::bid).tolist();
// 根据b表id 查询b表数据 并映射成map
Map<Long,B> bMap = bDao.findAll(bids).stream().toMap(b::id,x->x);
// 组装数据
。。。。

这些逻辑随处可见，写多了就比较烦，因此，需要一款框架来帮助我实现这个逻辑。
调研了一下市面上的框架，发现有一款框架easy_trans能解决这个问题，但是，这个框架对于目前我来说，也有一些问题：
1、框架太重了，比如框架引入了redis做缓存，有些项目没有redis，启动需要手动排除
2、数据翻译的实体需要继承框架默认的接口，改造较多
3、翻译后的字段是生成在增加map字段里面的，和前端联调时不方便，因为用的是swagger生成文档的，希望翻译后的字段能在文档中展示
因为自己的翻译需求比较简单，因此，决定自己开发一款框架解决这些问题。

二、架构设计

参考了easy_trans的源码，并设想翻译框架和spring validation一样，通过自定义不同注解，拓展翻译功能。
架构如下：

因此，设计一个翻译核心注解

@Retention(RetentionPolicy.RUNTIME)
@Target({ElementType.FIELD, ElementType.ANNOTATION_TYPE})
public @interface Trans {

    /**
     * @return 需要翻译的字段
     */
    String trans() default "";

    /**
     * @return 提取的字段
     */
    String key() default "";

    /**
     * @return 翻译数据获取仓库
     */
    Class<? extends TransRepository> using();

}

三、优点

1、核心源码简单，仅几百行，无任何依赖项；
2、高度可拓展，拓展逻辑仅仅只需要实现TransRepository接口
3、支持数据库翻译、字典翻译、集合翻译、嵌套翻译等
4、并行翻译，翻译不同字段是并行翻译的，性能高

四、基本使用

maven引入

        <dependency>
            <groupId>io.github.orangewest</groupId>
            <artifactId>easy-trans-core</artifactId>
            <version>0.0.3</version>
        </dependency>

比如现在有一个老师的实体对象

@Data
@AllArgsConstructor
@NoArgsConstructor
public class TeacherDto {

    private Long id;

    private String name;

    // 关联教哪个学科
    private Long subjectId;

}

课程科目实体对象

@Data
@AllArgsConstructor
@NoArgsConstructor
public class SubjectDto {

    private Long id;

    private String name;

}

学生实体对象

@Data
public class UserDto {

    private Long id;

    private String name;

    private String sex;

    @DictTrans(trans = "sex", group = "sexDict")
    private String sexName;

    private String job;

    @DictTrans(trans = "job", group = "jobDict")
    private String jobName;

    // 关联老师id
    private Long teacherId;

    @Trans(trans = "teacherId", key = "name", using = TeacherTransRepository.class)
    private String teacherName;

    @Trans(trans = "teacherId", key = "subjectId", using = TeacherTransRepository.class)
    private Long subjectId;

    @Trans(trans = "subjectId", using = SubjectTransRepository.class, key = "name")
    private String subjectName;

    public UserDto(Long id, String name, Long teacherId, String sex, String job) {
        this.id = id;
        this.name = name;
        this.teacherId = teacherId;
        this.sex = sex;
        this.job = job;
    }
}

我们在teacherName 上增加@Trans的注解，
其中trans 指名需要翻译哪个字段，
key说明需要使用的是哪个字段，
using 说明的是使用哪个数据仓库获取数据
TeacherTransRepository 代码如下：

public class TeacherTransRepository implements TransRepository {

    @Override
    public Map<Object, Object> getTransValueMap(List<Object> transValues, Annotation transAnno) {
        return getTeachers().stream().filter(x -> transValues.contains(x.getId())).collect(Collectors.toMap(TeacherDto::getId, x -> x));
    }

    public List<TeacherDto> getTeachers() {
        List<TeacherDto> teachers = new ArrayList<>();
        teachers.add(new TeacherDto(1L, "老师1", 1L));
        teachers.add(new TeacherDto(2L, "老师2", 2L));
        teachers.add(new TeacherDto(3L, "老师3", 3L));
        teachers.add(new TeacherDto(4L, "老师4", 4L));
        return teachers;
    }

}

模拟根据id查询，获取指定id的数据。
SubjectTransRepository

public class SubjectTransRepository implements TransRepository {

    @Override
    public Map<Object, Object> getTransValueMap(List<Object> transValues, Annotation transAnno) {
        return getSubjects().stream().filter(x -> transValues.contains(x.getId())).collect(Collectors.toMap(SubjectDto::getId, x -> x));
    }
    
    public List<SubjectDto> getSubjects() {
        List<SubjectDto> subjects = new ArrayList<>();
        subjects.add(new SubjectDto(1L, "语文"));
        subjects.add(new SubjectDto(2L, "数学"));
        subjects.add(new SubjectDto(3L, "英语"));
        subjects.add(new SubjectDto(4L, "物理"));
        return subjects;
    }

}

注册翻译仓库：

        TransRepositoryFactory.register(new TeacherTransRepository());
        TransRepositoryFactory.register(new SubjectTransRepository());
        TransRepositoryFactory.register(new DictTransRepository(new DictLoader() {
            @Override
            public Map<String, String> loadDict(String dictGroup) {
                return dictMap().getOrDefault(dictGroup, new HashMap<>());
            }

            private Map<String, Map<String, String>> dictMap() {
                Map<String, Map<String, String>> map = new HashMap<>();
                map.put("sexDict", new HashMap<>());
                map.put("jobDict", new HashMap<>());
                map.get("sexDict").put("1", "男");
                map.get("sexDict").put("2", "女");
                map.get("jobDict").put("1", "学习委员");
                map.get("jobDict").put("2", "生活委员");
                map.get("jobDict").put("3", "宣传委员");
                map.get("jobDict").put("4", "班长");
                map.get("jobDict").put("5", "团支书");
                map.get("jobDict").put("6", "团长");
                return map;
            }

        }));

代码测试：

    @Test
    void trans() {
        UserDto userDto = new UserDto(1L, "张三", 2L, "1", "2");
        System.out.println("翻译前：" + userDto);
        transService.trans(userDto);
        System.out.println("翻译后：" + userDto);
        List<UserDto> userDtoList = new ArrayList<>();
        UserDto userDto2 = new UserDto(2L, "李四", 1L, "2", "1");
        UserDto userDto3 = new UserDto(3L, "王五", 2L, "1", "3");
        UserDto userDto4 = new UserDto(4L, "赵六", 3L, "2", "4");
        userDtoList.add(userDto4);
        userDtoList.add(userDto3);
        userDtoList.add(userDto2);
        System.out.println("翻译前：" + userDtoList);
        transService.trans(userDtoList);
        System.out.println("翻译后：" + userDtoList);

    }

结果输出

翻译前：UserDto(id=1, name=张三, sex=1, sexName=null, job=2, jobName=null, teacherId=2, teacherName=null, subjectId=null, subjectName=null)
翻译后：UserDto(id=1, name=张三, sex=1, sexName=男, job=2, jobName=生活委员, teacherId=2, teacherName=老师2, subjectId=2, subjectName=数学)
翻译前：[UserDto(id=4, name=赵六, sex=2, sexName=null, job=4, jobName=null, teacherId=3, teacherName=null, subjectId=null, subjectName=null), UserDto(id=3, name=王五, sex=1, sexName=null, job=3, jobName=null, teacherId=2, teacherName=null, subjectId=null, subjectName=null), UserDto(id=2, name=李四, sex=2, sexName=null, job=1, jobName=null, teacherId=1, teacherName=null, subjectId=null, subjectName=null)]
翻译后：[UserDto(id=4, name=赵六, sex=2, sexName=女, job=4, jobName=班长, teacherId=3, teacherName=老师3, subjectId=3, subjectName=英语), UserDto(id=3, name=王五, sex=1, sexName=男, job=3, jobName=宣传委员, teacherId=2, teacherName=老师2, subjectId=2, subjectName=数学), UserDto(id=2, name=李四, sex=2, sexName=女, job=1, jobName=学习委员, teacherId=1, teacherName=老师1, subjectId=1, subjectName=语文)]

五、高级功能

1、自定义注解

使用@Trans注解标注在自定义注解上即可，自定义注解中需要有trans方法。
示例：

@Retention(RetentionPolicy.RUNTIME)
@Target({ElementType.FIELD})
// 使用@Trans标注
@Trans(using = TeacherTransRepository.class)
public @interface TeacherTrans {

    /**
     * 需要翻译的字段
     */
    String trans() default "";

    /**
     * key 提取的字段
     */
    String key() default "";

}

@Data
public class UserDto2 {

    private Long id;

    private String name;

    private List<Long> teacherIds;

    private List<String> jobIds;

    @DictTrans(trans = "jobIds", group = "jobDict")
    private List<String> jobNames;

    @TeacherTrans(trans = "teacherIds", key = "name")
    private List<String> teacherName;

    @TeacherTrans(trans = "teacherIds", key = "subjectId")
    private List<Long> subjectIds;

    @Trans(using = SubjectTransRepository.class, trans = "subjectIds", key = "name")
    private List<String> subjectNames;

    public UserDto2(Long id, String name, List<Long> teacherIds, List<String> jobIds) {
        this.id = id;
        this.name = name;
        this.teacherIds = teacherIds;
        this.jobIds = jobIds;
    }
}

测试

    @Test
    void trans2() {
        List<Long> teacherIds = new ArrayList<>();
        teacherIds.add(1L);
        teacherIds.add(2L);
        List<String> jobIds = new ArrayList<>();
        jobIds.add("1");
        jobIds.add("2");
        UserDto2 userDto = new UserDto2(1L, "张三", teacherIds, jobIds);
        System.out.println("翻译前：" + userDto);
        transService.trans(userDto);
        System.out.println("翻译后：" + userDto);
        List<UserDto2> userDtoList = new ArrayList<>();
        UserDto2 userDto2 = new UserDto2(2L, "李四", teacherIds, jobIds);
        List<Long> teacherIds2 = new ArrayList<>();
        teacherIds2.add(3L);
        teacherIds2.add(4L);
        List<String> jobIds2 = new ArrayList<>();
        jobIds2.add("3");
        jobIds2.add("4");
        UserDto2 userDto3 = new UserDto2(3L, "王五", teacherIds2, jobIds2);
        UserDto2 userDto4 = new UserDto2(4L, "赵六", teacherIds2, jobIds2);
        userDtoList.add(userDto4);
        userDtoList.add(userDto3);
        userDtoList.add(userDto2);
        System.out.println("翻译前：" + userDtoList);
        transService.trans(userDtoList);
        System.out.println("翻译后：" + userDtoList);
    }

结果输出

翻译前：UserDto2(id=1, name=张三, teacherIds=[1, 2], jobIds=[1, 2], jobNames=null, teacherName=null, subjectIds=null, subjectNames=null)
翻译后：UserDto2(id=1, name=张三, teacherIds=[1, 2], jobIds=[1, 2], jobNames=[学习委员, 生活委员], teacherName=[老师1, 老师2], subjectIds=[1, 2], subjectNames=[语文, 数学])
翻译前：[UserDto2(id=4, name=赵六, teacherIds=[3, 4], jobIds=[3, 4], jobNames=null, teacherName=null, subjectIds=null, subjectNames=null), UserDto2(id=3, name=王五, teacherIds=[3, 4], jobIds=[3, 4], jobNames=null, teacherName=null, subjectIds=null, subjectNames=null), UserDto2(id=2, name=李四, teacherIds=[1, 2], jobIds=[1, 2], jobNames=null, teacherName=null, subjectIds=null, subjectNames=null)]
翻译后：[UserDto2(id=4, name=赵六, teacherIds=[3, 4], jobIds=[3, 4], jobNames=[宣传委员, 班长], teacherName=[老师3, 老师4], subjectIds=[3, 4], subjectNames=[英语, 物理]), UserDto2(id=3, name=王五, teacherIds=[3, 4], jobIds=[3, 4], jobNames=[宣传委员, 班长], teacherName=[老师3, 老师4], subjectIds=[3, 4], subjectNames=[英语, 物理]), UserDto2(id=2, name=李四, teacherIds=[1, 2], jobIds=[1, 2], jobNames=[学习委员, 生活委员], teacherName=[老师1, 老师2], subjectIds=[1, 2], subjectNames=[语文, 数学])]

2、值提取

有些翻译类型匹配结果的时候是通过值去匹配的，比如说字典翻译，我们需要根据字典值去字典组里面去匹配数据，框架里面只需要标记key的值用#val即可实现，比如框架自带的@DictTrans注解。

@Trans(using = DictTransRepository.class, key = "#val")
@Retention(RetentionPolicy.RUNTIME)
@Target(ElementType.FIELD)
public @interface DictTrans {


    /**
     * @return 需要翻译的字段
     */
    String trans();

    /**
     * 字典组
     *
     * @return 字典分组
     */
    String group();

}

3、包装类翻译

有些类是包装类，比如返回的结果，返回的分页对象等，需要翻译的数据一般都是里面的实际业务对象，这时候，需要我们去配置解析包装类的解析器。
示例：

@Data
@AllArgsConstructor
public class Result<T> {

    private T data;

    private String message;

}

配置解析器，实现TransObjResolver接口即可

public class ResultResolver implements TransObjResolver {
    @Override
    public boolean support(Object obj) {
        return obj instanceof Result;
    }

    @Override
    public Object resolveTransObj(Object obj) {
        return ((Result<?>) obj).getData();
    }

}

TransObjResolverFactory.register(new ResultResolver());

测试：

    @Test
    void trans3() {
        List<Long> teacherIds = new ArrayList<>();
        teacherIds.add(1L);
        teacherIds.add(2L);
        List<String> jobIds = new ArrayList<>();
        jobIds.add("1");
        jobIds.add("2");
        jobIds.add("3");
        UserDto2 userDto = new UserDto2(1L, "张三", teacherIds, jobIds);
        Result<UserDto2> result = new Result<>(userDto, "success");
        System.out.println("翻译前：" + result);
        transService.trans(result);
        System.out.println("翻译后：" + result);
        UserDto2 userDto2 = new UserDto2(2L, "李四", teacherIds, jobIds);
        Result<UserDto2> result2 = new Result<>(userDto2, "success");
        Result<Result<UserDto2>> result3 = new Result<>(result2, "success");
        System.out.println("翻译前：" + result3);
        transService.trans(result3);
        System.out.println("翻译后：" + result3);
    }

结果输出：

翻译前：Result(data=UserDto2(id=1, name=张三, teacherIds=[1, 2], jobIds=[1, 2, 3], jobNames=null, teacherName=null, subjectIds=null, subjectNames=null), message=success)
翻译后：Result(data=UserDto2(id=1, name=张三, teacherIds=[1, 2], jobIds=[1, 2, 3], jobNames=[学习委员, 生活委员, 宣传委员], teacherName=[老师1, 老师2], subjectIds=[1, 2], subjectNames=[语文, 数学]), message=success)
翻译前：Result(data=Result(data=UserDto2(id=2, name=李四, teacherIds=[1, 2], jobIds=[1, 2, 3], jobNames=null, teacherName=null, subjectIds=null, subjectNames=null), message=success), message=success)
翻译后：Result(data=Result(data=UserDto2(id=2, name=李四, teacherIds=[1, 2], jobIds=[1, 2, 3], jobNames=[学习委员, 生活委员, 宣传委员], teacherName=[老师1, 老师2], subjectIds=[1, 2], subjectNames=[语文, 数学]), message=success), message=success)

4、与springboot集成

maven 引入

		<dependency>
			<groupId>io.github.orangewest</groupId>
			<artifactId>easy-trans-spring-start</artifactId>
			<version>0.0.3</version>
		</dependency>

翻译仓库实现TransRepository，翻译解析器实现TransObjResolver，并在实现类上标注@Component；
在需要翻译的对象属性上面标注好相关注解；
需要翻译方法上使用@AutoTrans注解，框架会自动拦截需要翻译的对象，实现翻译。

    @GetMapping("/query")
    @AutoTrans
    public Result<PageData<BizDTO>> page(Query query) {
        PageData<BizDTO> page = bizService.page(query);

        return new Result<PageData<BizDTO>>().ok(page);
    }

5、与orm框架集成

通常很多翻译需求，都是根据id去查询实体对象，我们可以通过orm框架对此进行统一翻译
示例：
与mybatis-plus集成
定义翻译注解

@Retention(RetentionPolicy.RUNTIME)
@Target({ElementType.FIELD})
@Trans(using = DbTransRepository.class)
public @interface DbTrans {

    String trans();

    String key() default "";

    /**
     * 数据库目标class
     */
    Class<? extends BaseEntity> entity() default BaseEntity.class;

}

翻译仓库实现

@Component
public class DbTransRepository implements TransRepository {

    @Resource
    private TransDriver transDriver;

    @Override
    public Map<Object, Object> getTransValueMap(List<Object> transValues, Annotation transAnno) {
        if (transAnno instanceof DbTrans) {
            DbTrans dbTrans = (DbTrans) transAnno;
            List<? extends BaseEntity> entities = transDriver.findByIds(getIds(transValues), dbTrans.entity());
            return entities.stream().collect(Collectors.toMap(BaseEntity::getId, x -> x));
        }
        return Collections.emptyMap();
    }

    /**
     * 获取查询id
     */
    private List<Long> getIds(List<Object> transValues) {
        return transValues.stream()
                .map(x -> (Long) x)
                .collect(Collectors.toList());
    }

}

public interface TransDriver {

    /**
     * 根据ids获取集合
     *
     * @param ids         ids
     * @param targetClass 目标类类名
     */
    List<? extends BaseEntity> findByIds(List<? extends Serializable> ids, Class<? extends BaseEntity> targetClass);

}

@Component
public class MybatisTransDriver implements TransDriver {

    @Override
    public List<? extends BaseEntity> findByIds(List<? extends Serializable> ids, Class<? extends BaseEntity> targetClass) {
        try (SqlSession sqlSession = SqlHelper.sqlSession(targetClass)) {
            BaseMapper<? extends BaseEntity> mapper = SqlHelper.getMapper(targetClass, sqlSession);
            return mapper.selectBatchIds(ids);
        }
    }

}

六、开源地址

https://github.com/orangewest/easy-trans
希望大家能给给小心心